Kerensky, Vasil M.

The purpose of this study was to investigate the relationship between mobility and academic performance among African-American fifth grade urban public school students. Blacks in South Florida are a diverse group. As of 1990 the U.S. Census Bureau identified 12 different ethnic groups that comprise the Black ethnic category in Broward County, Florida. The term African-American and the data used in this study reflect the ethnic diversity. This study examined Stanford Achievement Test scores in mathematics (SATM) and reading (SATR) over the years 1992, 1993 and 1994. The SATM (N = 2208) and SATR (N = 2178) scores over the three year period were the dependent variables in the study. Gender and mobility were the independent variables of the study. Mobility was divided into three categories of non-mobile, mobile and very mobile. Descriptive statistical procedures were applied to the data. The mean and standard deviations were established for the dependent variables for each year of the study. Analysis of variance was applied to the dependent and independent variables of the study over the three year period of 1992, 1993 and 1994. There were no statistically significant main effects or interactions (p >.05 for all effects). Consequently none of the null hypotheses were rejected. This study supports the conclusion that gender and mobility had no relationship with academic performance in urban African-American fifth grade public school students. Suggestions for further research include: (a) exploration of the linkage of gender and mobility with the variables of socioeconomic status, (b) family structure and character qualities in relationship to academic achievement in urban African-Americans, and (c) the replication of this study in a rural setting with students of a different race.

The major purpose of this study was to determine if the attire of academic advisors would significantly affect students' satisfaction with their academic advisors. Academic advisors wore three types of attire: professional, casual, and a special uniform. The major concern was to determine if the student's first impression, as influenced by attire, of the academic advisor was significant in terms of satisfaction with the advising process. Data collected from 363 community college students were analyzed using analysis of variance (ANOVA). Based upon the statistical analyses, the researcher failed to reject all the null hypotheses. The attire of the academic advisors did not significantly affect the students' rating of satisfaction as measured by the sum of scores on the Student Evaluation of Counselor/Academic Advisor Form. There were no significant interactions between attire and age, attire and gender, and attire and race/ethnic origin. The major conclusion that can be drawn from this study is that student satisfaction with academic advisors is not significantly affected by the academic advisor's attire. Neither age, gender nor race/ethnic origin of the community college student had any impact or interaction with attire on satisfaction with the academic advisor. The literature review made a compelling statement that academic advisors do make a difference in retention of students. Future research could determine how the attire (casual, professional or uniform) of the academic advisor impacts the productivity of the institution, student satisfaction and how the dress of the academic advisor impacts the academic advisors' view of their competence, comfort and rapport with the student. Use of a semantic differential survey instrument could be employed. Perception of academic advisor's attire may be more complex than previously thought. The information obtained from this study may be more useful as stimuli to encourage further investigation.

The purpose of this study was to investigate the justification of using Full Scale IQ scores as the principal means of determining whether adjudicated youths are as capable of engaging in higher-order thinking processes as youths in public schools. This study examined the performance of a group of adjudicated youths (N = 50) in a correctional training school and a comparative group in a public school district (N = 50) on 4 of 10 mandatory subtests which comprise the Wechsler Intelligence Scale for Children (WISC-R, WISC-III). The criteria for the subjective selection of subtests to be examined focused on choosing those which measured cognitive abilities and were not heavily influenced by acquired academic skills and socioeconomic factors. For each of the selected subtests, an inferential statistical analysis was made at a.01 level of significance using independent sample two-tailed t tests to compare the difference between the means of subtest scaled scores for subjects from the correctional training school and the public school district. The results of this analysis indicated that there was no significant difference between the cognitive abilities of the youths in the correctional training school and a comparative group in a public school district as measured by the selected subtests. This study supports the conclusion that the use of a Full Scale IQ score from a standardized intelligence test may not accurately reflect the individual cognitive abilities of adjudicated youth to engage in higher-order thinking processes. Recommendations were made for correctional educators to identify approaches which could be used that better target the higher-order thinking skills of an adjudicated youth population.

Ubiquitous incorporation of computer-based, electronic technologies into American society has created intense pressure upon the composition instructors at Florida's community colleges to develop computer skills themselves and to develop curricula which routinely incorporate these technologies into their teaching responsibilities. This fact makes clear the need to assess instructors' readiness: What computer skills have they acquired and what computer-aided teaching methodologies have they learned which make them capable of and willing to teach composition with the aid of computers? What support services on community college campuses seem to encourage instructors to take on the responsibility to teach computer-aided composition. After its pilot testing, a thirteen question survey was sent to the 473 known instructors of credit-level composition at Florida's twenty-eight community colleges. This study examined the characteristics of the 222 respondents, of whom 61 reported that they do teach computer-aided composition and 161 reported that they do not. One way analyses of variance applied to the survey responses of these groups compared them regarding their institutions' technology support services; regarding their familiarity with, training for, and uses of computer technologies in their personal and professional responsibilities; and regarding their participation in departmental, divisional, and institutional planning activities for technology enhancements. Findings revealed that, compared to those instructors who do not teach computer-aided composition, instructors who do are (1) more likely to be familiar with and use a greater variety of computer technologies for both personal and professional use, (2) more likely to serve on departmental committees which consider technology uses, (3) more likely to have in-office access to desktop computers, to have nearby full-service faculty/student computer labs, and to have greater access to computer hardware overall, and (4) more likely to have attended workshops and taken courses on computer technologies and on instructor methods for use of technologies. In responses to a Likert-like question, composition instructors who do teach with the aid of computer technology also expressed greater than expected professional satisfaction with computer-aided composition as compared to traditional methodology.

The United States Bureau of the Census has released much of the 1990 census data through CD-ROM technology. The census data is readily accessible, easily obtained, simple to use, and virtually free. This study presented an innovative way to use census data coupled with business and educational practices as a marketing and planning aid for adult literacy education. The purpose of this study was to develop a model that utilized geo-demographic census data to identify and segment the adult low-literate population. The identification and segmentation of the adult low-literate population gives literacy educators the ability to plan literacy related educational programs and to market literacy education programs to a well-defined, well-understood population. Demographic data from the 1990 United States census was compiled for the number of adults age 25 and older living in each census tract of Broward County, Florida, and St. Louis County, Minnesota, who possessed low-literacy identifiers found by past national literacy surveys. School districts from the two study counties supplied records containing the locations and the ABE and adult ESOL programs offered at the adult literacy education sites. Combining the census findings with the school district records revealed several combinations of conclusions. Some geographic locations with high incidences of the low-literacy identifiers, as could be expected, contained adult literacy education sites. However, some geographic locations with high incidences of the low-literacy identifiers did not contain adult literacy education sites. In contrast, some areas with low incidences of the low-literacy identifiers contained several adult literacy education sites. A Pearson correlation involving the low-literacy identifiers found a positive relationship between each of the educational attainment variables, each of the language isolation variables and between the educational attainment variables and the language isolation variables. In Broward County, Florida, there was a moderate negative relationship between per capita income and the other low-literacy identifiers. In St. Louis County, Minnesota, there was a weak relationship between per capita income and the other low-literacy identifiers. A K-means cluster analysis of census tracts found nine homogeneous clusters in Broward County, Florida, and seven homogeneous clusters in St. Louis County, Minnesota. The model could be a beneficial planning and marketing aid for adult literacy programs in any geographic location. Included in this study are 40 data tables. Census tract maps and thematic maps also illustrate the findings.

This study used a discriminant analysis procedure to identify the
demographic variables (school, teacher, grade level, subject, test used
for evaluation, semester of treatment, and gender) or subsets of variables
that would predict elementary students' success with computer assisted
instruction (CAl). Also, multiple factorial analyses of variance were
performed to test the interaction effect between treatment (CAl) and the
demographic variables. Recommendations suggest that future research attempt to identify
the appropriate teacher training in the use of computers for instruction that
will produce recurring student achievement with CAl. It is also suggested
that future research examining the academic effects of CAl in the
elementary classroom (grade 2 through 5) should not be concerned with
the grade level of the student, the subject being studied (mathematics or
language arts), the test used for evaluation (local or standardized), or the
gender of the student.

The purpose of this study was to examine the relationships between reading gains and the following demographic variables: economic status, ethnicity, and gender. The reading gains of limited English proficient (LEP) and English-fluent students in the school district of Palm Beach County, the fourth largest school district in Florida, were examined for two consecutive years. The researcher also sought to determine if LEP status affected the relationship between reading gains and the variables identified. The sample included 51,590 English-fluent and 11,907 LEP students in the first year. In the second year of the study, the sample included 53,400 English-fluent and 12,830 LEP students. Multiple regression analysis was used to test the correlations and significance of the predictor variables for the subgroups, LEP and English-fluent students. Factorial analysis of variance was used to determine if there was an interaction between LEP status and the regression models for each of the predictor variables. The results indicated that yearly reading gains of English-fluent and Limited English Proficient students in Palm Beach County Schools could not be predicted from the predictor variables used during the years of the study. The results of multiple regressions revealed that a very small percentage of the criterion variable could be explained by the predictor variables (R 2 values ranged from .000 to .003). In the cases where the regression model was statistically significant, the results were not practically significant but the product of a trivial relationship between the variables, perhaps due to the size of the sample. After examining the interaction of LEP status with each of the predictor variables, it was concluded that the interaction between LEP status and each of the predictor variables did not increase significantly the predictive value of the models. No significant difference was found in the regression slopes between LEP and English-fluent students for any of the predictor variables.

Since the passage of Public Law 94-142 (The Education of All Handicapped Children Act) which is currently called the Individuals with Disabilities Act (IDEA), there has been a movement to service students with disabilities in more inclusive environments. The Broward County School District had been involved in the State of Florida's effort to reform the special education funding model and the delivery of services to students in more inclusive environments since 1994. The purpose of this study was to determine the effects of three delivery models, resource room, co-teaching, and support facilitation, on the reading and math achievement of students with mild to moderate disabilities. The subjects of this study consisted of 231 students with mild to moderate disabilities enrolled in 29 schools and serviced in one of these three delivery models. Their achievement was measured during the 1998--1999 school year based upon the results of a nationally normed achievement test. An Analysis of Variance (ANOVA) as well as an Analysis of Covariance (ANCOVA) was performed using the spring of 1998 test scores as baseline data or covariates and the spring of 1999 scores as dependent or criterion variables. The independent variable was the type of delivery model. The results did not reveal any statistically significant differences in the achievement of the students in any of the three models thus indicating that in this study, the type of model did not have a significant impact on the achievement of these students. Implications based upon limitations as well as recommendations for further study are presented.

National Board for Professional Teaching Standards Certification is one way in which teachers can demonstrate excellence in teaching. This study was conducted in order to examine the effect of the professional development experiences on overall scores on the assessment of candidates for National Board Certification RTM in Florida. The study was conducted using the entire population of candidates in Florida (1,787) during the 2000--2001 assessment cycle. A survey was used to collect data from the candidates. Of the surveys sent, 62% were returned and used in the study. The survey included questions regarding the educational background, demographics, and professional development experiences of the teachers. To examine the effect of professional development experiences on the overall National Board Certification assessment scores, a multiple regression analysis was used. A predictive discriminant analysis was used to predict passing or non-passing group membership. The criterion variable used was the score on the National Board Certification assessment. Results indicated that professional development experiences coupled with background and demographics contributed significantly to a candidate's overall score. In the full model, controlling for all of the other variables in the model, holding an advanced degree in the field was found to contribute to the predictive accuracy of the model. Also, the number of hours preparing the portfolio and preparing for the assessment center also contributed to the predictive accuracy of the model. Bivariate correlations indicated that there was a positive correlation between candidates' overall scores and the number of hours of professional development. There was a positive correlation between the overall score and the amount of time preparing the portfolio entries and for the assessment center. The predictive accuracy of the full model for predicting passing and non-passing group membership was 58.7%. Tests of significance for the unique contribution of each subset of variables to the cross-validated classification accuracy of the full model were insignificant.

The Broward County, Florida school district has implemented an intern principal program for training new principals. One area of concern regarding the intern principal program is the process used for selecting mentors for intern principals. Toward generating a model for choosing mentors, the purpose of this study was to examine the relationship between six clusters of principal competencies and five clusters of mentoring traits. The relationship between the competencies and mentoring traits was examined to determine the degree to which it is possible to predict each mentoring trait cluster from ratings on the six principal competency clusters. The sample for this study was 91 principals in the Broward County, Florida public schools. The instrument used was the Mentor Identification Instrument which asked the sample to rate someone who had significantly influenced their career on the principal competencies and the mentoring traits. A Likert type scale was used. Canonical correlation indicated a significantly high degree of multivariate overlap between the principal competencies and the mentoring traits. Multiple regression showed that the principal competencies were highly predictive of each mentoring trait cluster. Further analysis yielded models that classified mentors and non-mentors with more than 80% cross-validated accuracy. Recommendations regarding the implementation of these selection models are included.